Figure 1: Measurements of total volatile organic compounds in five new houses in Pennsylvania, Michigan, and Florida and median concentration in U.S. EPA study.

In 1993, the Indoor Environment Program began investigating indoor air quality in new energy-efficient houses. Five new houses have been included in the study, all in the eastern U.S. Two had nearly identical floor plans and were part of a demonstration project near Pittsburgh, PA; one was built conventionally, while the other incorporated a number of energy-efficient features. The conventional house was studied for one year following construction, and the energy-efficient house was sampled on three occasions over a two-year period. The other three demonstration houses were in separate projects. One of these houses was investigated over a period of four months following construction. Ventilation rates ranged from 0.07 to 0.4 h^-1 (air changes per hour) with a median value close to 0.2 h^-1, indicating that these houses have tighter envelopes than older housing and that current building practices have improved building tightness.

Concentrations of airborne organic contaminants were measured in all of the houses. These measurements included formaldehyde, individual volatile organic compounds (VOCs), and total VOCs (TVOC). Since ventilation rate data were also collected, source strengths were calculated with a mass-balance model. These estimates of source strengths made it possible to compare the magnitudes of the emission sources in houses with different volumes and ventilation rates.

The concentrations of formaldehyde in all five houses were generally about 0.05 ppm (0.06 mg/m³), or lower, a level at which less than one percent of the population is expected to experience sensory irritation. The major sources of formaldehyde in houses are primarily particle board underlayment for floors and particle and fiberboard used for cabinetry and furniture. In an early-1980s study of formaldehyde concentrations in new energy-efficient houses, concentrations of 0.2 ppm were common. In the intervening years, composite-wood manufacturers have reduced the formaldehyde emissions of their products in response to concerns about the adverse health effects of this compound.

The concentrations of TVOC in the houses ranged from 0.65 to 12 mg/m³; the median value was 2.4 mg/m³ (see Figure 1). A large probability-based study of existing residences by the U.S. EPA found that the median concentration of TVOC representative of a population of about 600,000 was 0.7 mg/m³. Thus, the TVOC concentrations in the new houses were, with one exception, elevated compared to typical values. Calculated whole-house source strengths ranged from about 0.2 to 3 grams per hour of contaminants. Interestingly, there was only about a factor of two reduction in source strengths over the one- and two-year study periods in the two Pittsburgh houses. If interior finish materials such as paints and floor coverings are the dominant sources, source strengths should decrease substantially over a period of only a few months. Because of the magnitude and persistence of the source strengths in the Pittsburgh houses, the dominant source was probably not these finish materials but rather some other element of the houses (see below).

The dominant classes of compounds detected in the house air samples were terpene hydrocarbons, aromatic hydrocarbons, ketones, and aldehydes. In the Pittsburgh houses, the most abundant compounds were alpha-pinene, d-limonene, acetone, pentanal, hexanal, heptanal, nonanal and acetic acid. The aldehydes are of particular concern because they are irritants and produce unpleasant odors at low concentrations. Hexanal, the aldehyde with the highest concentrations and source strengths in the houses, ranged in concentrations from 0.02 to 0.9 mg/m³. Thus, concentrations of hexanal alone often significantly exceeded the odor threshold of 0.06 mg/m³. In the Pittsburgh houses, acetic acid also exceeded its odor threshold and probably accounted for some of the detectable odor.

The sources of the VOC emissions in these new houses are still under investigation through field and laboratory studies. One possibility is that structural components consisting of natural lumber, laminated lumber, plywood, and oriented strandboard are the dominant sources. These wood products are the most logical sources of terpene hydrocarbons, aldehydes, and acetic acid. Because they are present in a house in large amounts, they may form a sizable chemical reservoir that could be a source of contaminant emissions over extended periods. These contaminants may be entering the indoor air through walls, ceilings, and floors for duct work, plumbing, and electrical systems.

As demonstrated by the paired Pittsburgh houses, the problems of persistently elevated VOC concentrations and odors are probably not unique to energy-efficient houses, but rather may be relatively common in new houses built with current construction materials and techniques. Further research should be able to demonstrate ways to reduce emissions, thereby reducing concentrations of VOC contaminants in houses, by modifying materials and construction practices used for new houses, including those built with specific energy-efficient features. This solution might avoid the need for increased ventilation rates, which would diminish the energy-efficiency and comfort benefits of houses with tightly sealed envelopes.

--Alfred T. Hodgson

Alfred T. Hodgson
Indoor Environment Program
(510) 486-5301; (510) 486-6658 fax

This research is supported by DOE's Office of Building Technologies, and State and Community Programs.